Light field geometry of a Standard Plenoptic Camera

Area of Science:

• Optics and Photonics
• Computer Vision
• Computational Imaging

Background:

• Standard Plenoptic Cameras (SPC) capture multi-focal images from a single exposure.
• Conventional cameras lack the ability to derive depth information directly from a single shot.

Purpose of the Study:

• To develop a novel method for estimating distance and depth from Standard Plenoptic Camera (SPC) images.
• To derive the position and baseline of virtual lenses equivalent to a camera array.

Main Methods:

• Developed a ray tracing model based on paraxial approximation and linear equations.
• Implemented the model in Matlab and validated using Zemax optics simulation software.
• Designed a realistic SPC for experimental verification.

Main Results:

• Experimental results showed a prediction error of less than 11% for image refocusing distance at 3.5 m compared to Zemax simulations.
• Baseline estimations derived from the method showed no significant difference from simulation.
• The proposed methodology successfully estimates depth and distance parameters.

Conclusions:

• The developed methodology provides a viable approach for depth and distance estimation using SPCs.
• This technique offers a potential alternative to traditional depth map acquisition methods like disparity analysis.
• The research validates the accuracy of the proposed ray tracing model and estimations.